hello-kanister

Trivial examples of application-level backup and restore using Kanister blueprints

Developing

Fork and clone this repository, then navigate to the project root and follow the instructions below.

Install pre-commit hook (optional)

The pre-commit hook runs formatting and sanity checks such as tofu fmt to reduce the chance of accidentally submitting badly formatted code that would fail CI.

ln -s ../../hooks/pre-commit ./.git/hooks/pre-commit

Prerequisites

1. An AWS account
2. Valid AWS credentials for an IAM administrator account - see Configure the AWS CLI for details
3. OpenTofu v1.7.2 or later
4. A Kubernetes cluster - spin one up with kind if in doubt
5. Cluster-admin access to the Kubernetes cluster with kubectl
6. Helm
7. The Kanister operator installed on our Kubernetes cluster

Create the S3 bucket for storing our Kanister backups

Create the S3 bucket for storing our Kanister backups using OpenTofu:

tofu init
tofu plan
tofu apply

Supported variables:

Name	Type	Required	Default value	Description
profile	string	-	"default"	The profile to assume from the AWS credentials file
region	string	-	"ap-east-1"	The AWS region to assume

Running the time logger example

The time logger example is based on the Kanister tutorial.

First create a deployment which appends a timestamp to the log file at /var/log/time.log every second:

kubectl create -f time-logger/deployment.yaml

Wait for the pods in our deployment to become ready:

kubectl wait \
    --for=condition=Ready \
    pods \
    -l app=time-logger \
    --timeout=180s

Sample output:

pod/time-logger-777955ffb4-q6bmp condition met

The log file is streamed to STDOUT via the busybox container:

kubectl logs -l app=time-logger -c busybox

Sample output:

Sat Jun  8 13:56:27 UTC 2024
Sat Jun  8 13:56:28 UTC 2024
Sat Jun  8 13:56:29 UTC 2024
Sat Jun  8 13:56:30 UTC 2024
Sat Jun  8 13:56:31 UTC 2024
Sat Jun  8 13:56:32 UTC 2024
Sat Jun  8 13:56:33 UTC 2024
Sat Jun  8 13:56:34 UTC 2024
Sat Jun  8 13:56:35 UTC 2024
Sat Jun  8 13:56:36 UTC 2024

Create the ConfigMap and Secret which contain our bucket URL and AWS credentials for accessing our bucket:

kubectl create -f time-logger/configmap.yaml
kubectl create -f time-logger/secret.yaml

Now create our Kanister blueprint. The blueprint defines backup and restore actions for our log file but does not execute them.

kubectl create -f time-logger/blueprint.yaml

To run the backup action defined in our blueprint, create an ActionSet invoking that action.

kubectl create -f time-logger/backup-actionset.yaml

Wait a few seconds and confirm that the backup ActionSet has entered a "complete" state:

kubectl -n kanister get actionsets.cr.kanister.io

Sample output:

NAME                          PROGRESS   SIZEDOWNLOADEDB   SIZEUPLOADEDB   ESTIMATEDDOWNLOADSIZEB   ESTIMATEDUPLOADSIZEB   RUNNING PHASE   LAST TRANSITION TIME   STATE
time-logger-s3backup-n6h8q    100                                                                                                          2024-06-08T13:51:24Z   complete

Confirm that a log file time-logger/XXXX-XX-XX was uploaded to S3 as well.

Optionally run the restore ActionSet as well and confirm that it runs to completion, though you'll probably need to edit the restoreDate beforehand which is hardcoded to 2024-06-08 in the given template:

kubectl create -f time-logger/restore-actionset.yaml

Now view the contents of the log file /var/log/time.log within the pod and observe that log file was overwritten by the restore operation.

License

MIT